This is a repository copy of Compartmental Genomics in Living Cells Revealed by Single-Cell Nanobiopsy.

White Rose Research Online URL for this paper:http://eprints.whiterose.ac.uk/101074/

Version: Accepted Version

Article:

Actis, P orcid.org/0000-0002-7146-1854, Maalouf, MM, Kim, HJ et al. (4 more authors) (2014) Compartmental Genomics in Living Cells Revealed by Single-Cell Nanobiopsy. ACS Nano, 8 (1). pp. 546-553. ISSN 1936-0851

https://doi.org/10.1021/nn405097u

eprints@whiterose.ac.ukhttps://eprints.whiterose.ac.uk/

Reuse

Unless indicated otherwise, fulltext items are protected by copyright with all rights reserved. The copyright exception in section 29 of the Copyright, Designs and Patents Act 1988 allows the making of a single copy solely for the purpose of non-commercial research or private study within the limits of fair dealing. The publisher or other rights-holder may allow further reproduction and re-use of this version - refer to the White Rose Research Online record for this item. Where records identify the publisher as the copyright holder, users can verify any specific terms of use on the publisher’s website.

Takedown

If you consider content in White Rose Research Online to be in breach of UK law, please notify us by emailing eprints@whiterose.ac.uk including the URL of the record and the reason for the withdrawal request.

1

Compartmental Genomics in Living Cells Revealed by Single-cell

Nanobiopsy

Paolo Actis ¥, 1, Michelle M. Maalouf1, Hyunsung John Kim, Akshar Lohith, Boaz Vilozny, R.

Adam Seger, Nader Pourmand*

a Department of Biomolecular Engineering, University of California Santa Cruz, 1156 High

Street, Santa Cruz, CA 95064, USA

¥ Current address: Department of Medicine, Imperial College London, Du Cane Road, London

W12 0NN, United Kingdom

1 These authors contributed equally

*Corresponding Author: Department of Biomolecular Engineering, University of California

Santa Cruz, 1156 High Street, Santa Cruz, CA 95064, USA pourmand@soe.ucsc.edu, +1 (831)

502-7315

Keywords: nanopipette, scanning ion conductance microscopy, single-cell biopsy,

electrowetting, mithochondria, next-generation sequencing

2

Abstract

The ability to study the molecular biology of living single cells in heterogeneous cell populations

is essential for next generation analysis of cellular circuitry and function. Here, we developed a

single-cell nanobiopsy platform based on scanning ion conductance microscopy (SICM) for

continuous sampling of intracellular content from individual cells. The nanobiopsy platform uses

electrowetting within a nanopipette to extract cellular material from living cells with minimal

disruption of the cellular milieu. We demonstrate the subcellular resolution of the nanobiopsy

platform by isolating small subpopulations of mitochondria from single living cells, and quantify

mutant mitochondrial genomes in those single cells with high throughput sequencing

technology. These findings may provide the foundation for dynamic subcellular genomic

analysis.

3

Most tissues in the body consist of a fine architecture of cells with substructural mosaics of

cell clusters and intricate yet ill-defined circuits. Cells interact with other cells to produce precise

functioning of tissues, organs and organ systems. Understanding the molecular biology of

individual cells within circuits in complex environments is necessary to understand both normal

physiology and also how injury or disease causes pathology to alter cell function1.

Furthermore, investigations into the molecular basis of cancer cell transformation,

proliferation and metastasis would be aided by technologies capable of analysis of molecules

and genetic factors involved in disease progression at a single cell level. This is particularly true

for cancer because tumors are heterogeneous and consist of cells at different stages of

transformation. Studies on whole tissue provide only a statistical average of the many events

occurring in different cells. Single cell analysis may reveal how tumorigenesis-related signaling

pathways can cause genetic changes which ultimately lead to the transformation of a healthy

cell to a cancerous cell.

Anatomical approaches have been employed for a century to study tissue architecture and

cellular heterogeneity and have provided insights into cell-cell interaction and circuitry within

tissues. However, these approaches are usually employed post-mortem, reducing our ability to

study the dynamic function of individual cells in heterogeneous populations. Also, much of what

we know of individual cell molecular biology in our body is diluted by analysis of mixed cellular

populations in heterogeneous tissues.

Attempts to study single cell molecular biology in heterogeneous tissue cultures began

several decades ago when patch-clamp approaches were used to isolate RNA from neurons to

profile the gene expression of single cells2 linking genetics to neuronal function3. This approach

provided only a snapshot of single cell molecular biology, because once the neuron was

impaled by the micropipette it died. Over the last 20 years, other approaches have been

employed to study single cell molecular biology, including laser capture micro-dissection

techniques4. Again, these approaches either study cells postmortem, which prevents analysis of

dynamic changes in the cell’s genomics over time, or they do not maintain the natural

environment of a cell in heterogeneous tissues.

Nanoscale devices are ideal single-cell surgical tools because of their potential for high spatial

and temporal resolution studies5-8. Recently, two groups independently developed cellular

nanoendoscopes for single cell analysis. Singhal et al. 9 attached a carbon nanotube to the tip

of a glass micropipette and showed its potential for interrogating cells by transporting fluids and

performing optical and electrochemical diagnostics down to the single organelle level. Similarly,

4

Yan et al. 10 developed a nanowire waveguide attached to the tip of an optical fiber, which can

transmit visible light into intracellular compartments of a living mammalian cell and detect optical

signals from sub-cellular regions. The cylindrical shape of these nanoendoscopes allows the

probing of organelles deep inside the cell.

Integration of nanoendoscopes with scanning probe techniques may allow automation

and high–throughput analysis. In 2003, Osada and coworkers inserted atomic force microscope

(AFM) tips into living cells to extract mRNAs which were then analyzed with PCR techniques 11.

More recently, Wickramasinghe’s group optimized this method by coating AFM tips with

platinum, allowing extraction of the mRNA molecules through dielectrophoresis.

Wichramasinghe’s technology has been successfully combined with standard assay techniques

to detect RNA molecules in breast cancer cells 12, 13.

Scanning ion conductance microscopy (SICM) is a scanning probe technique that has

recently gained attention for its ability to image living cells with high spatial 14 and temporal

resolution 15. SICM relies on a glass nanopipette filled with an electrolyte solution biased to

generate an ion flow through the nanopipette aperture16, 17. The nanopipette is then scanned

over a surface, and changes in the magnitude of the measured ion current reflect the

topography of the sample 18, 19. Nanopipette have been recently utilized to measure the

electrophysiology at small synaptic boutons20, to locally deliver molecules to multiple subcellular

areas21, and to trap molecules in lipid bilayers22. Our group has developed nanopipettes as a

label-free sensing platform23-28 and we recently described an adaptation of SICM allowing the

multi-component injection of single cells in culture29. Laforge et al. 30 developed an

electrochemical attosyringe based on a glass nanopipette to deliver a minute amount of liquid

into a living cell via electrowetting. Briefly, when a nanopipette is filled with an organic solution

and immersed into an aqueous one, a liquid-liquid interface is formed at the nanopipette

opening. If a voltage is then applied across this interface, a force is generated that can induce

the aqueous solution to flow into/out of the nanopipette31.

To begin to study the molecular properties of living cells, we adapted an SICM platform that

uses electrowetting within a nanopipette to extract minute quantities of cellular material from

living cells in culture with minimal disruption and combine it with sensitive sequencing

technologies to study the genomics of individual cells and their organelles.

Nanopipettes were integrated into a custom-built Scanning Ion Conductance Microscope

(SICM) (Figure 1a) that allows automated positioning of the nanopipette nanometers above the

cell29. To adapt the SICM as a single-cell biopsy platform, the nanopipette was filled with a 10

mM THATPBCl solution in DCE, and fitted with a silver wire coated with AgTBACl (see online

5

methods). When a DCE-filled nanopipette is immersed into an aqueous solution, a liquid-liquid

interface is formed at the nanopore lumen due to the hydrophobic nature of DCE32, 33. The

application of a voltage across this interface induces a change in the DCE surface tension. This

effect, called electrowetting, causes the aqueous solution to flow in the nanopipette when a

negative voltage is applied and to flow out when the bias is reversed (Supplementary Fig. 2,

Supplementary Fig. 3, Video 1). From geometrical calculations this volume was estimated to be

~50fL, which corresponds to ~1% of the volume of a cell.

While in cell culture medium, the nanopipette is polarized with a positive bias to prevent

medium from flowing into the barrel. This bias generates an ion current through the liquid-liquid

interface which is used as the input into a feedback loop. Custom-designed software directs the

nanopipette toward the cell until it detects a 0.5% drop in the ionic current. At this point the

software stops the approach and quickly lowers the nanopipette by 1 たm at a high speed

(100たm/s) to pierce the cell membrane (Figure 1b), inserting the nanopipette tip into the cell

cytoplasm. The nanopipette bias is then switched to -500 mV for 5 seconds, which causes the

controlled influx of cell cytoplasm into the nanopipette (Figure 1b, c) followed by a switch to

100mV, which stops the influx but does not cause the efflux of aspirated contents. The

nanopipette is then quickly raised, and the aspirated content is transferred into a 5 uL droplet of

RNase free H2O by application of +1 V for 2 minutes and kept at 4°C.

Because of the small pore of the nanopipette (50 nm in radius), its insertion into cells is

minimally invasive (Figure 2a, Supporting Figure 1). This is a major advance over previous

approaches for single cell molecular analysis, which employed micropipettes that severely

damage cell membranes. As a result, the nanopipette technology can be used to sample living

cells multiple times in the life of a cell to study molecular dynamics. To show the effect of the

nanopipette on cell function is innocuous, human BJ fibroblast cells were loaded with the Ca2+

imaging agent Fluo4 AM and fluorescent microscopy was used to measure localized

intracellular [Ca2+] before, during, and after the nanobiopsy (Figure 1d). Optical micrographs

confirm that the procedure is minimally invasive, generating a barely detectable Ca2+ change

during nanobiopsy. The cell fully recovers within 5 seconds post aspiration, reaching [Ca2+] that

matches pre-aspiration level. In contrast, cellular biopsy using micropipettes employed for patch

clamp electrophysiology show dramatic changes in cytosolic Ca2+ concentration in the cell

(Supplementary Figure 7). Because the nanopipette is minimally invasive, it can be employed

for multiple sampling of cellular cytoplasm without overtly altering cell function. The 100-nm

diameter of the nanopipettes, limits the interaction of DCE with the cell membrane to an area of

~0.01 たm2. This area is about 0.004% of the total surface area of a cell in culture (assuming the

6

surface area of cell being equal to ~250 たm2).The disruption to the cell function is mainly caused

by the aspiration of cellular contents and the application of a potential though the cell

membrane. Furthermore, we performed cell viability studies that indicated a survival rate higher

than 70% for the 50+ nanobiopsies performed in this study.

Figure 1: Schematic of single cell nanobiopsy . a) Illustration of automated approach to cell

surface, penetration in the cell cytosol followed by controlled aspiration of cytoplasmic material

by electrowetting, and delivery of the biopsied material into a tube for analysis. Scheme not to

scale. b) Optical micrographs of the nanobiopsy procedure illustrating the approach, cell

membrane penetration, aspiration via electrowetting, and retraction of the nanopipette tip. Scale

bars 25たm. c) Chronoamperometry during nanobiopsy. A nanopipette filled with a 10mM

solution of THATPBCl penetrates a cell in culture. Before the penetration, the nanopipette is

biased at +100mV to prevent any aqueous solution from entering into the tip. After cell

membrane penetration, the bias is switched to -500mV. The entry of intracellular solution in the

nanopipette tip causes the ion current to increase. d) Calcium imaging during nanobiopsy

procedure showing false color fluorescent micrographs of human BJ fibroblast cells stained with

7

Fluo4 AM during nanobiopsy. Red arrows indicate where the nanopipette penetrated the cell.

Scale bars 25たm, unless otherwise specified.

To validate the nanobiopsy platform, small cytoplasmic samples were isolated from single

HeLa cells expressing GFP and PCR was used to amplify GFP transcripts (Figure 2a, 2b). The

total RNA was reverse-transcribed to generate cDNA and real time-PCR analysis of cDNA

confirmed the presence of GFP transcripts (see Online Methods). The real-time PCR

amplification plots show a Cq value of 20 for the positive control (total RNA extracted from

~1000 HeLa-GFP cells), 30 when the input RNA was the biopsied content of a single cell, and

33 for the negative control (water). The difference in Cq values from the positive control and

biopsied sample is 10, which corresponds to a 210-fold (1024 fold) difference in input material

(figure 2b).

8

Figure 2. RNA Analysis post nanobiopsy a) Scheme not to scale showing nanobiopsy

followed by RNA analysis and SEM micrograph of a representative nanopipette tip b) Post

biopsy analysis via qPCR targeting GFP RNA from HeLa cells showing a positive control of total

RNA from a ~1000 cells lysate (red curve, Cq 20), a nanobiopsy taken from within a single cell

9

(blue curve, Cq 30) and the negative control (black curve, Cq 33). c) Reproducibility of

nanobiopsy protocols. qPCR targeting GFP RNA. Positive control (red curve) consists of RNA

extracted for GFP HeLa cells spiked in the cDNA synthesis mix. Nanobiopsy (blue curve)

represents the average of 4 nanobiopsies with 4 different nanopipettes performed in a PBS

solution containing the lysate of ~10000 GFP HeLa cells following the protocol described in the

manuscript. Negative control (black curve) where no input material was used for cDNA

synthesis.d),e) Read coverage of RNA sequencing of aspirations are displayed in the UCSC

genome browser. Genomic position is displayed along the X axis and depth of coverage is

represented in the Y axis. d) Ensemble gene predictions are displayed in red, where lines

represent introns and rectangles represent exons. These gene predictions are on the reverse

strand and are displayed in the 3’ to 5’ direction. Alternative 3’ exon usage in the gene PABPC1

is seen in M3 and M7 vs M8. M8 shows coverage over the 3’ region of the longer isoform,

whereas M3 and M7 show coverage over the shorter isoform. e) The gene MCCC2 is displayed

in blue in the 5’ to 3’ direction. Read coverage extends across the entire length of the gene and

demonstrates the feasibility of generating full length cDNA from transcripts isolated by

nanobiopsy.

Because individual cells are known to have variability in gene expression, the reproducibility

of the nanobiopsy protocols was tested by performing aspiration in a solution containing the

lysate of ~10000 GFP HeLa cells. Nanobiopsy (Figure 2c, blue curve) represents the average

with error bars of 4 nanobiopsies performed with 4 different nanopipettes in a PBS solution

containing the lysate of ~10000 GFP HeLa cells. Post aspiration analysis via qPCR (Figure 2c)

confirmed the high reproducibility of the nanobiopsy protocols. Overall, out of the 50+

nanobiopsises performed in this study we obtained a success rate of 70% for RNA aspirations,

as confirmed by PCR analysis followed by gel electrophoresis (Supplementary Figure 6).

In addition to qPCR analysis, cDNA of mature mRNA was generated, sequenced and the

reads mapped to the human reference genome (see Online Methods). Differential 3’

untranslated region (UTR) usage in cytoplasmic Poly-A binding protein (PABPC1) was

observed, where two aspirations show dominant expression of a noncannonical isoform

whereas the third shows expression of the canonical isoform (figure 2d). We have also

observed instances where coverage was obtained over an entire transcript rather than just the

3’ UTR. An example of read coverage over the full length of methylcrotonoyl-CoA carboxylase 2

(MCCC2) as shown in Figure 2e34. These results demonstrate the feasibility of generating

mRNA expression from single living cells using the nanobiopsy platform. Finally, when enriched

10

for the most overrepresented Gene Ontologies, we found that two aspirations predominantly

expressed machinery related to translation, while one was enriched for processes related to

metabolism and morphogenesis demonstrating that nanobiopsy can be used to analyze

endogenous RNAs.

In addition to total RNA analysis of a cell, the nanobiopsy platform can be employed to

isolate and analyze cellular organelles, specifically mitochondria. Mitochondria are essential for

energy production and express their own genome distinct from the nuclear genome, providing a

way to establish the subcellular resolution of the nanopipette technique.

11

Figure 3. Mitochondrial nanobiopsy a) Fluorescent micrograph of human BJ fibroblast cells

stained with MitoTracker® Green before nanobiopsy (right panel) and after nanobiopsy (left

panel). Red circles highlight a dark spot caused by mitochondria removal. Scale bars 15 たm. b)

(left panel) Bright field image of the nanopipette tip (red circle) used for mitochondria

nanobiopsy in a). (Right panel) Negative fluorescent micrograph (black areas indicate high

fluorescence) of left panel showing fluorescence arising from the nanopipette tip which indicates

the success of mitochondria nanobiopsy. Scale bars 15 たm. c) Sequencing results demonstrate

variable conservation of heteroplasmic frequencies in aspirations. Heteroplasmic variants with

estimated frequencies between 5% and 99% are displayed as circles where the area of the

circle is proportional to the observed frequency. The nucleotide of the variant is specified by the

color, where ‘A’ is represented in red, ‘C’ is represented in violet, ‘G’ is represented in blue and

‘T’ is represented in green. The 14713 A->T variant at shows similar frequencies across

aspirations and population; whereas the 16278 C->T variant shows a greater variance of

heteroplasmic frequencies in aspirations. Low frequency variants were also found in both

aspirations, but not in the population.

To show that it was possible to isolate individual mitochondria from a cell and study the

mitochondria genome independently of the nuclear genome, human BJ fibroblasts were stained

using MitoTracker® Green (Invitrogen) (Figure 3a, Supplementary Figure 8), a fluorescent label

for mitochondria. The nanopipette was placed above a region of the cell with an abundance of

mitochondria, and a biopsy was taken (Figure 3a right panel). Fluorescence within the tip of the

nanopipette indicated the successful aspiration of mitochondria (Figure 3b right panel).

Next-generation sequencing was employed to sequence the genomes of the biopsied

mitochondria. The mitochondrial DNA was amplified using primers specific to the mitochondrial

genome. This approach is novel and important because it allows us to selectively sequence the

mitochondrial genome without the need to separate the mitochondrial nucleic acids from the

more abundant nuclear DNA. The amplified mitochondrion DNA was subjected to gel

electrophoresis which revealed that all 3 samples show bands at the correct length for human

mitochondrial DNA (Supplementary Figure 9).

The nanobiopsy relies on the voltage-controlled aspiration of cellular material and not on

adsorption to the nanopipette side walls. No amplification is observed if a negative voltage is not

applied to the nanopipette after insertion into the cell and when aspiration is performed in the

bulk solution (Supplementary Figure 4). This is the key element that differentiates the

nanobiopsy technology from AFM based platforms. Both Wickramasinghe’s and Osada’s group

12

used AFM probes to extract RNA from cells in culture, either based on physisorption or

hybridization of complementary RNA immobilized onto the probe11, 12. By contrast, the hollow

nature of nanopipette probes allows general fluid aspiration, thus its application is not limited to

nucleic acid.

Sequencing results from two mithocondria aspirations from a single cell, as well as DNA

extracted from a population of cells revealed that the frequencies of some heteroplasmic

variants are more conserved at the subcellular level (Figure 3c). For example, two

heteroplasmic variants (14713 A->T and 16278 C->T) were found both in subcellular aspirations

and samples from the total (pooled) mitochondrial population. Frequencies of the 14713 A->T

variant were found to be similar across all samples, 83%, 87%, and 79%. However, the 16278

C->T variant showed substantial variation in frequencies with 62%, 98% and 38% frequencies

for the pooled versus the selected mitochondria, respectively. Whereas the average

heteroplasmic frequency is observable in population-based sequencing, only subcellular

sequencing revealed heterogeneity of the variant within a cell. Based on genomic data of the

pooled mitochondrial population data alone, it would be impossible to know that the frequency of

the 16278 C->T variant fluctuated more than the 14713 A->T variant, despite similar

representation in the population. The 16278 C->T variant represents a common polymorphism

occurring in 1,845 out of 16,411 mitochondrial sequences in Genbank. Although this particular

variant does not appear to have any disease function, mitochondrial DNA mutations have been

implicated in tumorigenesis. Tracking the emergency of somatic variants could help elucidate

mechanisms by which cancers escape cell death and increase proliferation

In addition to the differences in heteroplasmic frequencies among subcellular aspirations,

we were also able to detect a number of variants that were > 5% of sequencing reads. We were

able to detect five variants between 5-15% in Aspiration 2 and four variants in Aspiration 1 (p-

value < 0.005 in both cases). These very rare variants were not detectable in the total

mitochondrial population sample and may represent variants within individual mitochondrial

genomes. These results show that combining the nanopipette with a novel mitochondrial

directed genomic technology, small subpopulations of mitochondria genomes in single cells can

be sequenced.

We have developed a minimally invasive nanosurgical technology to interrogate living cells.

Minute amounts of total RNA and mitochondrial DNA were taken out of a single cell and nucleic

acids were analyzed using next generation genomic sequencing techniques. Current methods

for single-cell manipulation often can detect only one class of analytes. For example, it is

13

possible to isolate total cellular RNA to conduct mRNA profiling in single cells. In contrast, the

nanobiospy platform has subcellular resolution and can be used to isolate small subpopulations

of mitochondria from single living cells, and quantify mutant mitochondrial genomes in those

single cells with high throughput DNA/RNA sequencing technology. As a consequence,

mitochondrial mutation rates can be assessed in single cells, a key step in interrogating why

and how cellular degenerative mutations gradually build up over time to cause cellular

dysfunction and death. The use of nanopipettes to aspirate just a few copies of mitochondrial

DNA from a living cell might provide the basis for less invasive and more accurate monitoring of

disease progression, or lay the foundation for developing new classes of drugs to attenuate

neurodegeneration, or to treat diseases as diverse as Parkinson's Disease, and Alzheimer’s

Disease and can be used as a platform in cancer research and clinical management, elucidating

the role of heterogeneity in primary tumor tissues and systemically identifying critical parameters

in disease progression and potential metastatic states35-38.

14

Materials and Methods

Electrowetting in nanopipettes

Nanopipettes were fabricated from quartz capillaries (Sutter Instrument, Novato, CA) using a P-

2000 laser puller (Sutter Instrument, Novato, CA). Mean diameter = (106 ± 16) nm

(Supplementary Fig. 1). Quartz nanopipettes were filled with a solution of 1-2 dichlorethane

(DCE) containing 10mM tetrahexylammonium tetrakis(4-chlorophenyl)borate (THATPBCl). A

silver wire coated with silver tetrakis(4-chlorophenyl)borate (AgTBACl) was then inserted into

the barrel of the nanopipette, while a Ag/AgCl wire was immersed in the bath solution acting as

a reference/counter electrode.

SICM setup

The scanning ion conductance microscope (SICM) consists of a Axopatch 200B low-noise

amplifier (Molecular Devices, Sunnyvale, CA) for nanopipette bias and current measurement, a

MP-285 micromanipulator (Sutter Instrument, Novato, CA) for coarse control in the X, Y, and Z

directions, a Nano-cube piezo actuator (Physik Instrumente, Irvine, CA) for fine control in the X,

Y, and Z directions, and a PCIe-7851R Field Programmable Gate Array (FPGA) (National

Instruments) for hardware control of the system. The system is operated using custom coded

software written in LabVIEW8.

Cell Culture

BJ Human Fibroblasts were purchased from Stemgent and cultured in Stromal Cell Basal

Medium (SCBM) (Lonza, Alpharetta, GA) with supplements (Lonza, Alpharetta, GA)with no

phenol-red. HeLa cells (Cell Biolabs, San Diego, CA) expressing Green Fluorescent Protein

were cultured in 70% DMEM medium supplemented with 10% fetal bovine serum, 0.1mM

essential amino acids, 2mM L-glutamine and 1% Pen-Strep as suggested from the supplier. All

cells were cultured in 5% CO2 at 37°C. Cells were plated on both ungridded and gridded plates

(ibidi µ-Dish35mm, highGrid-500 uncoated, sterile) coated with 1% gelatin. Cells were plated at

about 5 x 104 cells per dish.

cDNA synthesize and qPCR

cDNA was synthesized from total RNA nanobiopsied from cells using iScript™ cDNA Synthesis

Kit (BioRad, Hercules, CA) following manufacturer recommendations. Quantitative PCR (qPCR)

was used to confirm the success of the nanobiopsy protocols using Clontech’s Advantage

15

qPCR Premix SYBR Green I Master Mix (Clontech, Mountain View, CA) following manufacturer

recommendations. PCR primers: HeLa-GFP, Reverse primer:

GCGCCGAGGTGAAGTTCGAGG, Forward Primer: GCCGTCGCCGATGGGGGTGTT were

used to amplify the aspirated mitochondrial DNA to amplify 3968 base pairs visualized by gel

electrophoresis.

cDNA Synthesis and RNA Sequencing

Aspirated RNA samples were processed to cDNA using the Ovation® RNA-Seq system

(NuGEN Technologies, San Carlos, CA). The cDNA was prepared for individual aspiration for

library preparation. The quality and quantity of single-cell cDNA were evaluated using the

Agilent Bioanalyzer 2100 DNA High Sensitivity chip (Agilent, Palo Alto, CA).

For paired-end whole transcriptome library preparation, ~ 0.5-1.0 µg cDNA of each sample was

sheared to a size ranging between 200-300bp using the Covaris-S2 sonicator (Covaris,

Woburn, MA) according to the manufacturer’s recommended protocols. Fragmented cDNA

samples were used for the preparation of RNA-Seq libraries using TruSeq v1 Multiplex Sample

Preparation kit (Illumina, San Diego, CA). Briefly, cDNA fragments were end repaired, dA-tailed

and ligated to multiplex adapters according to manufacturer's instructions. After ligation, DNA

fragments smaller than 150bp were removed with AmPure XP beads (Beckman Coulter

Genomics, Danvers, MA). The purified adapter ligated products were enriched using

polymerase chain reaction (14 cycles). The final amplified libraries were resolved on 2.0%

agarose gel and size-selected in the range of 350–380bp using Caliper XT system

(PerkinElmer, Waltham, MA). The final RNA-Seq libraries were quantitated using the Agilent

bioanalyzer 2100 and pooled together in equal concentration for sequencing. The pooled

multiplexed libraries were sequenced, generated 2X150 bp paired-end reads on MiSeq

(Illumina, San Diego, CA).

Mitochondrial DNA amplification and Sequencing

Aspirated mitochondrial DNA was amplified by long and accurate PCR using a pair of primers;

forward TCA TTT TTA TTG CCA CAA CTA ACC TCC TCG GAC TC and reverse CGT GAT GTC

TTA TTT AAG GGG AAC GTG TGG GCT AT to generate a ~8K bp fragment. The aspirated

mitochondrial genomes were amplified using long and accurate polymerase chain reaction (LA-

PCR) LA PCR Kit Ver. 2.1 (Takara Bio). The mitochondrial genome amplification was carried out

in a 25たL reaction mixture containing aspirated template DNA, 2.5たL 10X Epicenter’s boost,

16

2.5たL 10X LA PCR buffer II (Mg2+ plus), 4.0たL dNTP (2.5 mM), 1.25たL of each primer (10たM),

0.125たL of LA Taq DNA Polymerase, and 11.4たL sterile distilled water. Thermocycler conditions

were 95°C for 2 minutes, followed by 30 cycles, each consisting of 94°C for 15 seconds then

68°C for 8 minutes, and a final extension at 68°C for 13 minutes. The mitochon drial amplicons

were sheared by Covaris S2 system resulting in 300-400bp fragments that were subjected to

automated Illumina multiplex paired end library preparation on robot. A unique index sequence

was used for each mitochondrial sample and sequenced in single Illumina HiSeq 2000 lane.

Bioinformatics Methods

Short reads from mitochondrial experiments were preprocessed to trim sequencing adapters

from the 3’ ends of reads. Mitochondrial reads were aligned to the revised Cambridge

Mitochondrial Reference Sequence using a nonspliced aligner. Heteroplasmic variants between

5% and 99% were reported. Short reads from RNA-sequencing experiments were preprocessed

to trim sequencing adapters from the 3’ ends of reads. Ten bases from the 5’ end of RNA-seq

reads to remove biases introduced during second strand synthesis of cDNA. RNA-seq reads

were aligned to the hg19 UCSC human reference genome34 and to UCSC known genes using a

spliced alignment tool. Genes were considered expressed if at least one read mapped uniquely

to an annotated transcript. Geneset enrichment analysis was performed on detected genes and

overrepresented Gene Ontologies were reported. Detailed bioinformatics methodology is

available in the Methodology section of the Supporting Information.

Acknowledgements: We thank Prof. Holger Schmidt and Dr. Tom Yuzvinsky of the Keck

Center for Nanoscale Optofluidics at UC Santa Cruz for electron microscopy of nanopipettes.

We thank Miten Jain and Brandon Rice for help in preliminary experiments. This work was

supported in part by grants from the National Cancer Institute [U54CA143803], and the National

Institutes of Health [P01-35HG000205].

17

References

1. Schubert, C. Single-cell analysis: The deepest differences. Nature 2011, 480, 133-137. 2. Van Gelder, R. N.; von Zastrow, M. E.; Yool, A.; Dement, W. C.; Barchas, J. D.; Eberwine, J. H. Amplified RNA synthesized from limited quantities of heterogeneous cDNA. Proceedings of the National Academy of Sciences 1990, 87, 1663-1667. 3. Eberwine, J.; Bartfai, T. Single cell transcriptomics of hypothalamic warm sensitive neurons that control core body temperature and fever response: Signaling asymmetry and an extension of chemical neuroanatomy. Pharmacol. Ther. 2011, 129, 241-259. 4. Emmert-Buck, M. R.; Bonner, R. F.; Smith, P. D.; Chuaqui, R. F.; Zhuang, Z.; Goldstein, S. R.; Weiss, R. A.; Liotta, L. A. Laser capture microdissection. Science 1996, 274, 998-1001. 5. Tian, B.; Cohen-Karni, T.; Qing, Q.; Duan, X.; Xie, P.; Lieber, C. M. Three-Dimensional, Flexible Nanoscale Field-Effect Transistors as Localized Bioprobes. Science 2010, 329, 830-834. 6. Yum, K.; Wang, N.; Yu, M.-F. Nanoneedle: A multifunctional tool for biological studies in living cells. Nanoscale 2010, 2, 363-372. 7. Meister, A.; Gabi, M.; Behr, P.; Studer, P.; Vrs, J. n.; Niedermann, P.; Bitterli, J.; Polesel-Maris, J. r. m.; Liley, M.; Heinzelmann, H.; Zambelli, T. FluidFM: Combining Atomic Force Microscopy and Nanofluidics in a Universal Liquid Delivery System for Single Cell Applications and Beyond. Nano Lett. 2009, 9, 2501-2507. 8. Zheng, X. T.; Li, C. M. Single living cell detection of telomerase over-expression for cancer detection by an optical fiber nanobiosensor. Biosens. Bioelectron. 2010, 25, 1548-1552. 9. Singhal, R.; Orynbayeva, Z.; Kalyana Sundaram, R. V.; Niu, J. J.; Bhattacharyya, S.; Vitol, E. A.; Schrlau, M. G.; Papazoglou, E. S.; Friedman, G.; Gogotsi, Y. Multifunctional carbon-nanotube cellular endoscopes. Nature Nanotech. 2011, 6, 57-64. 10. Yan, R.; Park, J.-H.; Choi, Y.; Heo, C.-J.; Yang, S.-M.; Lee, L. P.; Yang, P. Nanowire-based single-cell endoscopy. Nature Nanotech. 2012, 7, 191-196. 11. Uehara, H.; Kunitomi, Y.; Ikai, A.; Osada, T. mRNA detection of individual cells with the single cell nanoprobe method compared with in situ hybridization. Journal of Nanobiotechnology 2007, 5, 7. 12. Nawarathna, D.; Chang, R.; Nelson, E.; Wickramasinghe, H. K. Targeted messenger RNA profiling of transfected breast cancer gene in a living cell. Anal. Biochem. 2011, 408, 342-344. 13. Nawarathna, D.; Turan, T.; Wickramasinghe, H. K. Selective probing of mRNA expression levels within a living cell. Appl. Phys. Lett. 2009, 95, 083117. 14. Novak, P.; Li, C.; Shevchuk, A. I.; Stepanyan, R.; Caldwell, M.; Hughes, S.; Smart, T. G.; Gorelik, J.; Ostanin, V. P.; Lab, M. J.; Moss, G. W. J.; Frolenkov, G. I.; Klenerman, D.; Korchev, Y. E. Nanoscale live-cell imaging using hopping probe ion conductance microscopy. Nat Meth 2009, 6, 279-281. 15. Shevchuk, A. I.; Novak, P.; Taylor, M.; Diakonov, I. A.; Ziyadeh-Isleem, A.; Bitoun, M.; Guicheney, P.; Lab, M. J.; Gorelik, J.; Merrifield, C. J.; Klenerman, D.; Korchev, Y. E. An alternative mechanism of clathrin-coated pit closure revealed by ion conductance microscopy. The Journal of Cell Biology 2012, 197, 499-508. 16. Chen, C.-C.; Zhou, Y.; Baker, L. A. Scanning Ion Conductance Microscopy. Annual Review of Analytical Chemistry 2012, 5, 207-228. 17. Rheinlaender, J.; Geisse, N. A.; Proksch, R.; Schffer, T. E. Comparison of Scanning Ion Conductance Microscopy with Atomic Force Microscopy for Cell Imaging. Langmuir 2010, 27, 697-704. 18. Hansma, P.; Drake, B.; Marti, O.; Gould, S.; Prater, C. The scanning ion-conductance microscope. Science 1989, 243, 641-643.

18

19. Korchev, Y. E.; Bashford, C. L.; Milovanovic, M.; Vodyanoy, I.; Lab, M. J. Scanning ion conductance microscopy of living cells. Biophys. J. 1997, 73, 653-658. 20. Novak, P.; Gorelik, J.; Vivekananda, U.; Shevchuk, Andrew I.; Ermolyuk, Yaroslav S.; Bailey, Russell J.; Bushby, Andrew J.; Moss, Guy W. J.; Rusakov, Dmitri A.; Klenerman, D.; Kullmann, Dimitri M.; Volynski, Kirill E.; Korchev, Yuri E. Nanoscale-Targeted Patch-Clamp Recordings of Functional Presynaptic Ion Channels. Neuron 2013, 79, 1067-1077. 21. Babakinejad, B.; Jönsson, P.; López Córdoba, A.; Actis, P.; Novak, P.; Takahashi, Y.; Shevchuk, A.; Anand, U.; Anand, P.; Drews, A.; Ferrer-Montiel, A.; Klenerman, D.; Korchev, Y. E. Local Delivery of Molecules from a Nanopipette for Quantitative Receptor Mapping on Live Cells. Anal. Chem. 2013, 85, 9333-9342. 22. Jönsson, P.; McColl, J.; Clarke, R. W.; Ostanin, V. P.; Jönsson, B.; Klenerman, D. Hydrodynamic trapping of molecules in lipid bilayers. Proceedings of the National Academy of Sciences 2012, 109, 10328-10333. 23. Actis, P.; Rogers, A.; Nivala, J.; Vilozny, B.; Seger, R. A.; Jejelowo, O.; Pourmand, N. Reversible thrombin detection by aptamer functionalized STING sensors. Biosens. Bioelectron. 2011, 26, 4503-4507. 24. Umehara, S.; Karhanek, M.; Davis, R. W.; Pourmand, N. Label-free biosensing with functionalized nanopipette probes. Proceedings of the National Academy of Sciences 2009, 106, 4611-4616. 25. Actis, P.; Jejelowo, O.; Pourmand, N. Ultrasensitive mycotoxin detection by STING sensors. Biosens. Bioelectron. 2010, 26, 333-337. 26. Vilozny, B.; Wollenberg, A. L.; Actis, P.; Hwang, D.; Singaram, B.; Pourmand, N. Carbohydrate-actuated nanofluidic diode: switchable current rectification in a nanopipette. Nanoscale 2013, 5, 9214-9221. 27. Vilozny, B.; Actis, P.; Seger, R. A.; Vallmajo-Martin, Q.; Pourmand, N. Reversible Cation Response with a Protein-Modified Nanopipette. Anal. Chem. 2011, 83, 6121-6126. 28. Actis, P.; Vilozny, B.; Seger, R. A.; Li, X.; Jejelowo, O.; Rinaudo, M.; Pourmand, N. Voltage-Controlled Metal Binding on Polyelectrolyte-Functionalized Nanopores. Langmuir 2011, 27, 6528-6533. 29. Adam Seger, R.; Actis, P.; Penfold, C.; Maalouf, M.; Vilozny, B.; Pourmand, N. Voltage controlled nano-injection system for single-cell surgery. Nanoscale 2012, 4, 5843-5846. 30. Laforge, F. O.; Carpino, J.; Rotenberg, S. A.; Mirkin, M. V. Electrochemical attosyringe. Proceedings of the National Academy of Sciences 2007, 104, 11895-11900. 31. Dale, S. E. C.; Unwin, P. R. Polarised liquid/liquid micro-interfaces move during charge transfer. Electrochem. Commun. 2008, 10, 723-726. 32. Liu, S.; Li, Q.; Shao, Y. Electrochemistry at micro- and nanoscopic liquid/liquid interfaces. Chem. Soc. Rev. 2011, 40, 2236-2253. 33. Girault, H.; Schiffrin, D. Electrochemistry of liquid-liquid interfaces. Electroanalytical chemistry 1989, 15, 1-141. 34. Kent, W. J.; Sugnet, C. W.; Furey, T. S.; Roskin, K. M.; Pringle, T. H.; Zahler, A. M.; Haussler; David. The Human Genome Browser at UCSC. Genome Res. 2002, 12, 996-1006. 35. Clark, I. E.; Dodson, M. W.; Jiang, C.; Cao, J. H.; Huh, J. R.; Seol, J. H.; Yoo, S. J.; Hay, B. A.; Guo, M. Drosophila pink1 is required for mitochondrial function and interacts genetically with parkin. Nature 2006, 441, 1162-1166. 36. Kraytsberg, Y.; Kudryavtseva, E.; McKee, A. C.; Geula, C.; Kowall, N. W.; Khrapko, K. Mitochondrial DNA deletions are abundant and cause functional impairment in aged human substantia nigra neurons. Nat Genet 2006, 38, 518-520. 37. Chatterjee, A.; Mambo, E.; Sidransky, D. Mitochondrial DNA mutations in human cancer. Oncogene 0000, 25, 4663-4674. 38. Ståhlberg, A.; Thomsen, C.; Ruff, D.; Åman, P. Quantitative PCR Analysis of DNA, RNAs, and Proteins in the Same Single Cell. Clin. Chem. 2012, 58, 1682-1691.

19